> programming courses is that they tend to have a more-than-normal bimodal distribution of results There is argument that is not true [1]. However, I agree wit

> programming courses is that they tend to have a more-than-normal bimodal distribution of results There is argument that is not true [1]. However, I agree with your next statement that this isn't so much a 'lack of geek gene', but a lack of catching up. In [2], even something as simple as assignment operations were able to predict success in a CS course. I'm currently developing my own 'remedial tutorials' and researching the proper order of exercises to give a low-performing student. Similar to your experience, these help those low-performers catch up. [1] http://www.cs.toronto.edu/~sme/papers/2016/icer_2016_bimodal.pdf http://www.cs.toronto.edu/~sme/papers/2016/icer_2016_bimodal... [2] http://www.ppig.org/sites/default/files/2014-PPIG-25th-Ahadi-Lister_0.pdf http://www.ppig.org/sites/default/files/2014-PPIG-25th-Ahadi...